Abstract
Summary Microarray technology was used to characterize and compare hexachlorocyclohexane (HCH) contaminated soils from Spain. A library of 2290 hypervariable 16S rRNA gene sequences was prepared with serial analysis of ribosomal sequence tags (SARST) from a composite of contaminated and uncontaminated soils. By designing hybridization probes specific to the 100 most abundant ribosomal sequence tags (RSTs) in the composite library, the RST array was designed to be habitat‐specific and predicted to monitor the most abundant polymerase chain reaction (PCR)‐amplified phylotypes in the individual samples. The sensitivity and specificity of the RST array was tested with a series of pure culture‐specific probes and hybridized with labelled soil PCR products to generate hybridization patterns for each soil. Sequencing of prominent bands in denaturing gradient gel electrophoresis (DGGE) fingerprints derived from these soils provided a means by which we successfully confirmed the habitat‐specific array design and validated the bulk of the probe signals. Non‐metric multidimensional scaling revealed correlations between probe signals and soil physicochemical parameters. Among the strongest correlations to total HCH contamination were probe signals corresponding to unknown Gamma Proteobacteria , potential pollutant‐degrading phylotypes, and several organisms with acid‐tolerant phenotypes. The strongest correlations to alpha‐HCH were probe signals corresponding to the genus Sphingomonas , which contains known HCH degraders. This suggests that the population detected was enriched in situ by HCH contamination and may play a role in HCH degradation. Other environmental parameters were also likely instrumental in shaping community composition in these soils. The results highlight the power of habitat‐specific microarrays for comparing complex microbial communities.